MambaTrack: Exploiting Dual-Enhancement for Night UAV Tracking

TL;DR

This paper introduces MambaTrack, a novel night UAV tracking method that uses dual enhancement techniques and a cross-modal network to improve performance, efficiency, and robustness in low-light conditions.

Contribution

It presents a dual-enhancement tracker with a mamba-based low-light enhancer and a cross-modal network, significantly improving night UAV tracking performance and efficiency.

Findings

2.8× faster than CiteTracker

Reduces 50.2% GPU memory

Achieves advanced tracking performance in low-light conditions

Abstract

Night unmanned aerial vehicle (UAV) tracking is impeded by the challenges of poor illumination, with previous daylight-optimized methods demonstrating suboptimal performance in low-light conditions, limiting the utility of UAV applications. To this end, we propose an efficient mamba-based tracker, leveraging dual enhancement techniques to boost night UAV tracking. The mamba-based low-light enhancer, equipped with an illumination estimator and a damage restorer, achieves global image enhancement while preserving the details and structure of low-light images. Additionally, we advance a cross-modal mamba network to achieve efficient interactive learning between vision and language modalities. Extensive experiments showcase that our method achieves advanced performance and exhibits significantly improved computation and memory efficiency. For instance, our method is 2.8$\times$ faster than CiteTracker and reduces 50.2$\%$ GPU memory. Our codes are available at \url{https://github.com/983632847/Awesome-Multimodal-Object-Tracking}.